Job Stats Set For Shake-up

June 1, 2001 ( - Today's unemployment numbers could provide more questions than answers, as the Bureau of Labor Statistics unveils the impact of some structural changes in the compilation of the jobs data.

The changes include a change in the way companies are surveyed for the payrolls report and a new basis to adjust the numbers for seasonal fluctuations, as well as regular annual revisions to the figures.

Tis The Season

Get more!  Sign up for PLANSPONSOR newsletters.

Those seasonal adjustments could change the data going all the way back to 1996, and in the process could paint a very different picture of the job market over the past few months.

For example, the April unemployment report revealed an American economy cutting jobs at the fastest rate in a decade, some 223,000 jobs on top of 53,000 in March, according to Reuters. But research from UBS Warburg estimates that the new seasonal factors should lower the number of jobs lost in April by more than 50,000 while pulling down some of the strength at the beginning of the year.

The BLS, which released the new seasonal adjustment factors for payrolls last Friday, said that the seasonal adjustment factors built into the May data will amount to 728,000 jobs compared to 757,000 last year.

The May adjustment is an average adjustment for the year. The largest adjustment comes in January when factors like poor weather usually boost the number of layoffs. August is usually the month with the smallest adjustment.

“Small” Sampling

Stirring up more dust will be the so-called annual benchmark revisions- annual changes to estimates that the Labor Department makes each quarter to account for the number of jobs added from small and new businesses not covered by the survey.

The adjustment is made since the DOL cannot survey the small and new businesses every month but is able to estimate job growth in that sector based on tax and unemployment insurance records. The number will be adjusted to include state unemployment insurance that has actually been collected.

And if that weren’t enough, the May report will also feature the second phase of a three-phase overhaul of the sampling system, the way firms are selected for participation in the monthly payroll survey. This phase will extend the new system to manufacturing, construction and mining. Phase one, which included changes to the measurements of wholesale trade, was introduced a year ago.

May’s report is due to be released at 8:30 a.m. Economists surveyed by Reuters on average expect the number of workers on nonfarm payrolls to fall by 17,000, while the unemployment rate is expected to edge up to 4.6% from 4.5% in April.